I ran into a performance issue on multi-core architecture (8+) with software written in C ++ / VistualStudio / WindowsXP.
Suddenly, I realized that I had no idea about the characteristics of my L1 and L2 cache and the memory bandwidth of CPU-> to->.I tested several tools (including VTune, Glowcode, etc. etc.), but all of them fail when testing when loading in a multi-core architecture (for this reason I need them!).
Can you suggest any other tool that is not so suitable for executing graphs, but can give me at least some indications of cache / memory performance or can offer code fragments for the manual application of my application?
Thank!
source
share